| HOME | BIO | RESEARCH |
|
Welcome! This is the front page for Evan Oswald's thesis work. Thesis members are currently being invited to be part of the overseeing committee. Currently the thesis work regarding urban climate is tentatively done and has a paper published, the work concerning EHE trends project has the work for the publication completed and the last third (evaluating high resolution datasets) is in the early analysis stages. Below are links to the latest versions of the prospectus outline and manuscript. |
|
Thesis Goals The main subject of this thesis is elevated surface air temperatures and the motivation is that elevated temperatuers are a hazzard to the public. The following are specific goals of the proposed work: 1. Being relevant to the extreme temperature discussion from the end user's perspective 2. Further the understanding of temporal trends and spatial pattern of extreme temperature 3. Evaluate the ability of high resolution gridded datasets to describe extreme temperatures 4. Characterize the temperature at the local scale, and demonstrate the driving forces behind the variability. |
|
Being relevant to the end user Studying the most physically, or statistically, insightful climate variables is indispensable to futhering the understanding of the complex climate system. However, studying the variables, conditions, datasets or metrics specifically relevant to the end user scientist (public health, ecology, urban planning, etc.) is crucial to maximizing the benefit to the general public. The epidemiological literature has previously established relationships between temperature and mortality. Thus the relevance here comes from focusing on trends of, and how urban heat islands impact, those variables determined important to human mortality. This studies relevance also comes from the evaluation of datasets that have high potential to the end user scientist. These high resolution observational gridded datasets can be used to study EHEs (or other climate related variables such as fire risk or pathogen movement) directly or through data products that depended on them for creation. |
|
Studying trends of extreme heat There are many studies in the climatology literature about the past trends in surface temperatures. Trends in the daily mean or maximum/minimum temperatures are common and trends in the exceedence of extreme percentile thresholds are not especially rare either. However trends in EHEs (duration requirements) are rare, especially if the focus is on a certain area or time period(s). As this part of the thesis work, an analysis of the trends in EHE characteristics (intensity, timing within the heat season, number per season, duration) was undertaken. It focusd on the continental U.S. over a long period (81 years) and quantified the 1930-1970 (general cooling) and 1970-2010 (general warming) periods seperately. The differences between trends of EHEs with different requirements (e.g. the daily minimum, maximum or both daily extremes) were also quantified. Additionally, the spatial structure of the different trends was visually evaluated. |
|
Evaluation of high resolution observational datasets Recently, extremely high resolution surface climate observational datasets have become popular and readily available for use. Such spatial resolutions lend themselves to trend analyses at a level small enough to resolve a city, or downscaling a global climate model for predicative purposes. Additionally, the typical temporal resolutions allow for, amoungst other applications, hydrological modeling and EHE trend quantification. However, the appropriateness of such datasetsto describe trends have scarcely been quantified. As the second part of the thesis work, an evaluation of the ability of three high resolution gridded climate datasets to reproduce both trends in EHEs and single day extreme temperatures. This study will focus on the continental U.S. over the life of all datasets and compare them to the a reference dataset (USHCN). This study was designed to not evaluate the interpolation methods used to grid the datasets, but rather the affect of the lack of quality control employeed. The evaluation is designed to give confidence to anyone wanting to use these datasets for extreme temperature studies and possibly other studies. |
|
Characterization of local scale variability within a populated region Temperature variables relevant to heat stress vary not only from within to outside an urban region, but also within the urban regions themselves. The study of climate tendencies within these regions is interesting for those interested in the alteration of weather by physical landscape, however interest from the public health and urban planning scientists is also high. With new generations of reanalysis, model and even observational datasets now obtaining resolutions capable of distinguishing whole metropolitian regions (~5-15 km); the question of further downscaling to the intra-urban level (kilometers) is a logical final step. As the third part of the thesis work, a characterization of the spatial variability both in general and with respect to elevated tempertures was performed on the Detroit, Michigan, USA region. Moreover the magnitude of spatial variability was linked to the large-scale weather variables, both from observations and models. Lastly, the spatial variability of temperatures within a urban landscape was shown to be a response to the land cover/land use. Findings suggest that the spatial variability is stronger and more coherent during the daily low, than during the daily high. Findings that spatial variability exists during periods of high heat stress indicate that spatial variability within urban environments is likely important during EHEs. Additionally, the results indicated that local percent impervious surface and distance to sizeable water body were the most important drivers of the spatial variability. The majority of this work has been completed already and a manuscript (also reflecting the secondary authors and funding agencies goals) was recently published in the Journal of Applied Meteorology and Climate! |